Dry type transformer K-factor rating

[new_pe]

What references/resources would you recommend to determine whether a transformer should be specified with a k-factor rating?

Would the method you use differ between an existing installation and a new design?

On a new design, unless you were provided with exact equipment cut sheets with detailed electrical characteristics how could you make the determination of whether a k-factor rated transformer should be specified and what k-factor should be used? Doesn't some harmonic distortion cancel out among loads on a 3 phase system? Is it generally accepted practice to specify a k-factor transformer based on loads that have historically caused greater harmonic distortion, as most transformer manufacturers suggest? With newer less harmonic contribution power supplies being used in DC powered equipment, is k-factor still an issue?

Does the overall percentage of harmonic distorting loads vs total load on a transformer make a difference in considering specifying a transformer with a k-factor rating?

On an existing electrical system - should measurements be taken on the load that will be fed from the new transformer? What measurements would be most critical to determine whether a k-factor transformer should be specified? THD? % harmonic component at the odd harmonics? Voltage and current wave forms?

 

[jim dungar]

In my opinion, for the vast majority of situations (maybe >95%) there is absolutely no reason to use K-rated transformers.
For those few installations where the transformer L-N load will be >80% for more than 80% of its life, then K-rating may be worth investigating.

Have you ever heard of a utility installing a k-rated transformer?

 

[templdl]

I agree with Jim. Knowing that heat reduces transformer life expectancy, that harmonics cause heating, and that it is not that unusual for transformers to loaded anywhere near their capacity, it is unlikely that the life of standard transformer would be significantly reduced.
But first do a harmonic study to find out what they are then determine what you load actually will be. Then the hard part would be to stand behing the transformer that you detemine will do the job.
Also, even though 200% may be required, in theory I'm not so sure that current will exceed 100% realistically anyway. It would be interesting to put a clamp on ammeter on a neutral of a system with high harmonics.

But, that doesn't mean that the loading will not increase on the transformer in the future increasing the heating.

 

[T.M.Haja Sahib]

Doesn't some harmonic distortion cancel out among loads on a 3 phase system?

No.But in a 240v/120v single phase supply system,the third harmonics cancel out in the neutral.

 

[jim dungar]

Doesn't some harmonic distortion cancel out among loads on a 3 phase system?

Absolutely there may be some cancellation, this is why IEEE-519 is written to be applied at the Point of Common Coupling (PCC) and not at the line terminals of a device.

 

[jdsmith]

What references/resources would you recommend to determine whether a transformer should be specified with a k-factor rating?

I would suggest measuring the K factor of the loads that will be connected. K factor is a series summation of the THD% of all the harmonics - the higher harmonics are weighted to impact the K factor more than the lower harmonics because the higher the harmonic, the more heating the transformer experiences.

I actually measured some Dell servers and other process control equipment a while back and I measured a K value of about 1.4 - which will work perfectly fine on partially loaded standard (K=1) transformers.

Would the method you use differ between an existing installation and a new design?

An existing installation is easier to measure than a new design.

On a new design, unless you were provided with exact equipment cut sheets with detailed electrical characteristics how could you make the determination of whether a k-factor rated transformer should be specified and what k-factor should be used? Doesn't some harmonic distortion cancel out among loads on a 3 phase system? Is it generally accepted practice to specify a k-factor transformer based on loads that have historically caused greater harmonic distortion, as most transformer manufacturers suggest? With newer less harmonic contribution power supplies being used in DC powered equipment, is k-factor still an issue?

With modern switched power supplies, as opposed to early generation switched power supplies or linear power supplies, the THD and K factor are both low, similar to the 1.4 that I measured from a new Dell server with a modern, high PF, low harmonic switched mode power supply. Using the rules of thumb that the transformer manufacturers suggest may be appropriate for old utilization equipment with linear power supplies, but you're not likely to find that any more. Those guidelines are still published to help them sell K rated transformers.

Does the overall percentage of harmonic distorting loads vs total load on a transformer make a difference in considering specifying a transformer with a k-factor rating?

It does, but with modern power supplies and typical cushion that is put on the sizing of transformers it probably doesn't matter. It's all about heat. A transformer rated 75 kVA, 40 deg C ambient, 115 deg C rise is designed with enough copper and iron to perform suitably under those operating conditions with load at K=1. If the K factor of the load is higher those harmonics will generate extra heat in the transformer, so that same 75 kVA transformer could be operated successfully at lower loads as long as the neutral current doesn't get too high from the triplen harmonics (3rd and 9th) that will add up on the neutral. Don't worry about the neutral sizing in real life with modern power supplies - it used to be an issue with linear supplies years ago, and some marekting literature still proclaims this problem even though it's only applicable to older equipment.

On an existing electrical system - should measurements be taken on the load that will be fed from the new transformer? What measurements would be most critical to determine whether a k-factor transformer should be specified? THD? % harmonic component at the odd harmonics? Voltage and current wave forms?

If you were aware that the existing utilization equipment has very old power supplies it would be worth measuring K factor. If you don't have a meter that measures K factor then measure the %THD for each individual harmonic and calculate the K factor of the load yourself.